4 ms·
> "All three of these tests were null results. (i.e. No significant difference in aggregate purchases between either of the two options" (for goog&payp vs stri
by jacobn 14y ago
> "All three of these tests were null results. (i.e. No significant difference in aggregate purchases between either of the two options" (for goog&payp vs stripe, and several other combos)
This is a very interesting result. I would have thought that additional payment options would boost sales on the margin i.e. increase conversion by a couple of percent (not percentage points ;) for e.g. international customers, or people who already have accounts with GG/PP.
Patrick: Any chance you could elaborate a little more here? Would be very interesting to see some more detailed stats on this aspect.
- patio11 14y agoThanks for asking. So it turns out that two of them weren't null results, they were results against the change significant at 90% confidence. (Important difference for A/B testing practitioners, of only mild interest in context of the post.) I updated the post with the correction. If you really like stats, well, here you go: http://www.bingocardcreator.com/abingo/results http://www.bingocardcreator.com/abingo/results Rollover any conversion rate for raw stats. "Offer Credit Cards As Sole Checkout Option" tested Stripe vs. the three checkout methods. "Purchasing Page To Send People To"'s "Classic" is http://www.bingocardcreator.com/purchasing.htm http://www.bingocardcreator.com/purchasing.htm, "Cc Checkout" was the hacky Boostrap CC form. (Not currently available online.) "Offer Credit Cards As Checkout Option" was GC/PP/Stripe vs. GC/PP.